The Sure Things and The Hail Mary

Albert Einstein's photoelectric effect (PV) is the process by which the photons in sunlight are converted into electricity. It is the reverse process that is observed in familiar LED lights. The materials which exhibit this effect are small band gap semiconductors.

There are (currently) four major technical approaches to photovoltaic generation of electricity. Each has an proportional balance of advantages and risks. Each approach is a combination of light gathering (how much sunlight is collected per unit area of a cell), PV conversion efficiency (how much electricity is generated per area of sunlight) and cost (Dollars per Watt of energy produced by the cell).

Each is well represented in Silicon Valley today.


Crystalline Silicon (Ex.: SunPower)
This is a sure thing. Based on the same traditional crystalline silicon that is used in electronics, crystalline silicon solar cells have the advantage of being using extremely well understood physics as well as high volume manufacturing practices. There are literally tens of thousands of industry seasoned experts on this material in the electronics industry. This results in an early advantage in unit cost and manufacturing scalability. Crystalline silicon offers high photoelectric conversion efficiency, but is very complex and expensive to produce. Crystalline silicon cells are actually made from the "waste" wafers that are unusable for electronics.

However, Even with these seeming advantages, crystalline silicon is very complex to manufacture and after forty years of cost reduction, it is a fair concern that flat plate silicon solar cells should not expect many more major cost reductions. (or should we? watch for a future blog..)

The Bottom Line: This is what you see on rooftops today. Nearly all solar cells in existence to day are crystalline silicon based cells. Today they are less expensive than other exotic approaches, but it may be difficult to reduce cost further.

Light Gathering: 1 Sun
PV Conversion Efficiency: 15-18%
Current cost per Watt: $4.80


Amorphous Silicon Thin Film (Ex. Applied Materials)
Amorphous simply means non-crystalline. Also comprised of familiar silicon, this is a sure thing as well. Because it lacks the delicate crystal lattice structure of crystalline silicon, it is much less expensive to produce (imagine filling a sandbox with sand versus building a house of cards). For the same reason, its PV conversion efficiency is also lower -but not proportionally. Amorphous silicon based solar cells must be larger in order to generate the same amount of power, but the resulting cost will be lower. Amorphous silicon also benefits from incumbent status: It is the material used in LCD displays of which ten square miles will be produced in 2007.

However, while industry experience reducing the cost of amorphous silicon is not as mature as that of crystalline silicon, it has already seen cost fall by 20x. There is certainly still room for improvement, but the bottom of the well may not be far off.

The Bottom Line: Amorphous silicon is slightly less efficient at converting sunlight into electricity but much less expensive to produce. While the physics and manufacturing issues surrounding amorphous silicon are well enough understood that it can be produced in volume today, it is not so well understood that we should not expect further breakthroughs and cost efficiencies. All of the World's expertise with crystalline silicon for electronics can easily be ported to amorphous silicon.

Light Gathering: 1 Sun
PV Conversion Efficiency: 8-12%
Current cost per Watt: $3.99


Concentrated Photovoltaic (Ex.: SolFocus, Silicon Valley Solar)
This starts to be more of a long shot. The Concentrated PV approach addresses the high cost of the semiconductor material by using optics to focus a given area of light onto a small unit area of extremely high efficiency , but still proportionally less expensive compound semiconductor material. For example, by using 500x optical concentration (500 Suns), concentrated PV cell will generate as much electricity with 1 cm2 of semiconductor material as a standard (1 Sun) PV cell would produce with 500 cm2 of semiconductor. The idea is that the reduced cost of semiconductor material required will more than offset the added cost of optics and tracking motors (CPV cells typically must be aligned with precision toward the sun, tracking it across the sky through the day and the seasons)

However, it remains to be seen if the structural and maintenance costs of optics and motors over a twenty year lifetime outdoors will be justified by the accompanying increase in efficiency.

The Bottom Line: The principal of trading mystically delicate and expensive semiconductor material for apparently simple optics and tracking motors is reasonable, but there is a real challenge to the reliability and maintenance costs of anything that sits on a roof for twenty years. No one has been able to make this quite work yet.

Light Gathering: 2 to 500 Suns
PV Conversion Efficiency: 26%
Projected cost per Watt: <$3.00


CIGS Thin Film (Miasole, Nanosolar)
CIGS (Copper, Indium, Gallium, Selenide) is the Hail Mary, the long shot that could win the game. The preceding approaches all struggle with the manufacturing cost of their PV semiconductor material. CIGS approach seeks to reduce the cost of the semiconductor itself. Discovered in 1975, CIGS is a relatively new material, lacking the benefit of an enormous knowledge base of silicon. Despite this humble beginning, CIGS has demonstrated PV efficiency as high as 20% and thanks to a manufacturing process that requires comparatively moderate temperature and pressure control relative to silicon, it promises a dramatic reduction in energy cost. As an added benefit, CIGS PV cells are flexible, allowing cosmetic applications that are not practical with brittle crystalline silicon cells. It also greatly simplifies the demands of the panel structure -further reducing cost.

However, put simply, no one has ever been able to produce CIGS solar cells in high volume despite more than $140M of venture capital invested in Silicon Valley alone. In order to succeed, CIGS companies must develop new material physics, new manufacturing techniques and combine them into high reliability, high volume manufacturing processes. The risk is high, but the reward is its match.

The Bottom Line: CIGS has the best chance of delivering solar electricity at an price that will compete with coal. If CIGS is successful in its ambitions, it will be most important commercial development of this century.

Light Gathering: 1 Sun
PV Conversion Efficiency: up to 20%
Projected cost per Watt: <$1.00


What comes next? There's a very good bet. Watch for a future installment of The Solar Evangelist...